NEIGHBOR PROXIMITY ANALYSIS, A NEW TECHNIQUE FOR ANALYZING BIOTIC INTERACTIONS IN THE FOSSIL RECORD

Fine-scale interactions of organisms with their physical and biotic environment are important influences on community structure and diversity. Because of concerns over taphonomic overprinting, interactions such as competition, predation, mutualism, and microhabitat partition, seldom have been the focus of paleontological research. We present a method, neighbor proximity analysis (NPA), for identifying potential fine-scale interactions in settings in which the degree of time-averaging and transport is not readily apparent. The approach consists of (a) high-resolution mapping of individual organisms on a bedding plane; (b) generation of random assemblages in space through resampling of the observed data; (c) comparison of the extent of conspecific clustering and isolation of individuals between the random and observed assemblages. As time-averaging will tend to distribute individuals randomly on a bedding surface, conspecific clustering and isolation of individuals in space should become more rare with increased time/exposure. Therefore, non-random distributions suggest possible in situ preservation and may provide insight into community composition prior to taphonomic effects. This method is illustrated using fossiliferous horizons of the Rapid Member of the Little Cedar Formation (Givetian), which is beautifully exposed at the Devonian Fossil Gorge in Coralville, Iowa. The beds were deposited in open-marine, subtidal conditions and contain two recurring communities, one commonly associated with packstones, and the other with wackestones. Neighbor proximity analysis of the packstone community reveals that the brachiopod *Spinatrypa bellula* clusters with conspecifics more than would be expected based on random simulations (p < 0.02). Studies in the Recent suggest that conspecific clustering of sessile, benthic epifauna may (a) result from interspecific competition; (b) reduce predation risk; (c) reduce transport risk, or (d) indicate a favorable microhabitat. This example demonstrates that NPA not only makes it possible to identify potential interactions, but also can be a useful tool for assessing time-averaging.